#Q1 #Graph(I) 1) The first plot wants to answer the question of the percentage of flights,departing from Denver airport, delayed 15 minutes and more. 2)To our opinion the the plot shows the information well. The colors of each delay category shows to which destination the proportion of delayed flights was high/low. Which can bevery useful for optimizing scheduling. 3) The main question that arises is what could be the reason for all these delays. It would be useful to check the connection to the airline carrier or to see if there are spesific times that have more delays. 4)We believe that the graph would be easier to understand if the lines on thegraphwould thinner. It wouldn’t have blocked such a large area of the map,blocking the many state abbreviations.

#Graph (II) 1) The graph shows the cycles of the number of flights per day, and the number of delayed flights from all flights every day.The graph answer the question among all the departing flight how many of the departing flights are with a delay of at least 15 minutes.

  1. Yes we can see clearly each day how many out of the total flights had a delay. since both total flights and delayed flights are potted together it gives us a good picture of how many flights are delayed with respect to all flights in general. 3)Yes. IF we take a look at the graph from 31 August to 2 November, we can see that departing flights are steady at around 700 per day.On the other hand we can see the graph of the delayed flights does fluctuate quite. Perhaps the amount of departing flights each day does not effect the amount of delayed flights. 4)We believe the information would be better understood if the plot would be split in to two separate plots. it is simply easier on the eyes.

#Q2 #Graph of USA map with with connecting routs showing the percetage of departure from JFK airport.

  1. #graphic summarizing the flight volume and fights delayed,by date and showing weekly cycles.

#Q3

We can see from both plots how many flights in total each had and the percentage of flights with arrival delays of more then 15 minutes. Plotting these graphstogether gives us a good picture on how well each airline does with minimizing delayed flights. For instance we can see that UA has the most flights in total but does not have the highest percentage of flights being delayed. Hence we can say that that amount of flights are not necessarily connected to high percentage of delays.

In this graph we Try to understand which variables related to weather effect the departure delay and how they effect each other. We focused on aircraft with up to 100 seats. we choose the weather variable : humid,temperature,wind speed , viability (how far we can see on miles). The Statistical measurement we work with are join distribution correlation matrix and approximate the histogram of the data.

we can see we have strong a strong connection between arrivals delay and departure delay the correlation there is 0.94 also the join distribution graph looks like line and not like random cloud.

Also we can see some little connection between humid and arrivals delay the correlation there is 0.24 and in the join distribution graph we can that most of the point found mostly in the left area of the graph.

In addition There is An interesting connection between visib and humid the correlation there is -0.54. Also the joint distribution graph looks like A straight line.

Finally the wind_speed is distrusted like Normal distribution And departure delay,arrivals delay came from the same distribution family with different parameters.

#Q4 We will check the null hypothesis:The proportion of delayed flights per month is independent across months. First we plot the real data in a bar pot. Then we will simulate plots from the real data, and try to identify the real data in the lineup. WE will plot 19 simulated plots to get a significance level of 95%. If we can detect the true data then we will reject the null hypothesis

#Lineup when examining the lineup it is hard to identify the real data from the simulated data, with out having prior knowledge of the true data. Thus we can not reject the null hypothesis i.e we can’t say that the proportion of delayed flights per month is independent across months. ** The true data is in plot 1 **